On the rb9 the rear wing endplate extensions seem much more twisted than on the other cars. The axis of the vortex they interact with should pass along the center of them , since the twist is given to them to have a corrcect angle f attck with respect to the vortex induced perturbation of the local velocity angle.
From the orientation of the twist you can infer the sens of rotation of the vortex. To me, for the right side of the car seen froma the back, it seems that the vortex rotates counterclockwise - that seems strange because the lower endplate vortex is clockwise there